[1214] Mast Cells as a Part of a Prognostic Panel in Hodgkin Lymphoma
MT Howard, B Pohlman, P Elson, ED Hsi. Cleveland Clinic, Cleveland
Background: Therapy for classical Hodgkin Lymphoma (cHL) therapy results in cure in a most patients, but a subset of cHL patients suffer relapse and die from their disease. Biologic prognostic markers may assist in risk stratification of cHL patients. MAL and BCL2 expression in Reed-Sterberg (RS) cells and T-cell subsets (FOXP3/Granzyme B(GzB) ratio) have been shown to be prognostic markers in cHL. Other microenvironmental constituents such as mast cells (MC) have been suggested as additional markers. We studied our cohort of patients for MC infiltrates and correlated it with outcome.
Design: A tissue microarray was constructed using 99 patients with cHL. Immunohistochemistry (IHC) for tryptase was performed on these microarrays. MCs were counted and expressed as MC/hpf (using a 40x objective and 20mm ocular). The MCs/hpf (hot spot counting) were compared to the available clinical data, failure free survival (FFS) and overall survival (OS).
Results: The 99 patients had a mean age of 30 years, and 50/99 were male. Most of the patients (62%) had Stage I/II disease, International Prognostic Score (IPS) values 
2 (80%), and nodular sclerosis histology (82%). The mean MC hotspot count was 5.2 (range 0- 38.6). In 28% of cases, the RS cells expressed Bcl-2, 19% expressed MAL, and 75% of patients had FoxP3+:GzB+ ratios >1.0 (i.e. a FoxP3 
bias
). Multivariable analysis considered age, stage, IPS, histology, presence of bulky disease, MC count, FoxP3 bias, and expression of Bcl-2 and MAL.
| Factor | Hazard Ratio | P-value |
| MAL (pos./neg.) | 5.14 | <.001 |
| FoxP3 Bias (present/absent) | 5.84 | .002 |
| Bcl-2 (pos./neg.) | 5.68 | .009 |
| MC count (>6/6) | 3.56 | .01 |
Age ( 45/<45) | 3.06 | .02 |
Using the hazard ratios from the final models in Table 1, a scoring system was created that counts the number of poor prognostic factors present. Patients were stratified into two groups: patients that had 0-1 poor prognostic factors and patients that had two or more. Estimated 5-year FFS for patients with 0-1 factors was 89%, vs. 51% for patients with 2 or more factors (P<.001). Estimated 5-year OS for patients with 0 or 1 factors was 100%, vs. 63% for patients with 2 or more factors (P<.001).
Conclusions: Increased MCs is associated with a poorer prognosis in patients with cHL. As part of a battery of immunophenotypic findings evaluating both tumor cells and the microenvironment, MC counts may be useful in risk stratification for cHL.
Category: Hematopathology
Wednesday, March 11, 2009 9:30 AM
Poster Session V # 195, Wednesday Morning